EP0587450B1 - Verfahren und Anordnung zur Bildverarbeitung - Google Patents

Verfahren und Anordnung zur Bildverarbeitung Download PDF

Info

Publication number
EP0587450B1
EP0587450B1 EP93307181A EP93307181A EP0587450B1 EP 0587450 B1 EP0587450 B1 EP 0587450B1 EP 93307181 A EP93307181 A EP 93307181A EP 93307181 A EP93307181 A EP 93307181A EP 0587450 B1 EP0587450 B1 EP 0587450B1
Authority
EP
European Patent Office
Prior art keywords
area
rectangle
image
attributes
areas
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Lifetime
Application number
EP93307181A
Other languages
English (en)
French (fr)
Other versions
EP0587450A1 (de
Inventor
Kazuyuki C/O Canon Kabushiki Kaisha Saito
Tohru c/o Canon Kabushiki Kaisha Niki
Tadanori C/O Canon Kabushiki Kaisha Nakatsuka
Akihiro C/O Canon Kabushiki Kaisha Matsumoto
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Canon Inc
Original Assignee
Canon Inc
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority claimed from JP4243252A external-priority patent/JPH0696275A/ja
Priority claimed from JP4256315A external-priority patent/JPH06110994A/ja
Priority claimed from JP5002700A external-priority patent/JPH06208625A/ja
Application filed by Canon Inc filed Critical Canon Inc
Publication of EP0587450A1 publication Critical patent/EP0587450A1/de
Application granted granted Critical
Publication of EP0587450B1 publication Critical patent/EP0587450B1/de
Anticipated expiration legal-status Critical
Expired - Lifetime legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/60—Analysis of geometric attributes
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/11—Region-based segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
    • G06V30/40—Document-oriented image-based pattern recognition
    • G06V30/41—Analysis of document content
    • G06V30/413—Classification of content, e.g. text, photographs or tables
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10004—Still image; Photographic image
    • G06T2207/10008—Still image; Photographic image from scanner, fax or copier
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20036—Morphological image processing
    • G06T2207/20044—Skeletonization; Medial axis transform
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30176—Document

Definitions

  • the invention relates to an image processing apparatus for, particularly, dividing an input image into a character area and an area of a figure, table, or the like in an electronic apparatus such as OCR (optical character recognizing) apparatus, copying apparatus, facsimile apparatus, or the like.
  • OCR optical character recognizing
  • a conventional image processing apparatus there is an apparatus using a run length analyzing method as area dividing means whereby a distribution of run lengths is examined and a character area, a figure area, or the like is divided in accordance with a length of white run or black run, an apparatus using a spectrum analyzing method whereby a Fourier spectrum of an input image is analyzed and the input image is divided into various kinds of areas, an apparatus using a projection analyzing method whereby projections (histograms) in the vertical and horizontal directions are alternately repeatedly obtained and the area is divided from the information of a peripheral portion as disclosed in JP-A-1-15889.
  • a run length analyzing method as area dividing means whereby a distribution of run lengths is examined and a character area, a figure area, or the like is divided in accordance with a length of white run or black run
  • an apparatus using a spectrum analyzing method whereby a Fourier spectrum of an input image is analyzed and the input image is divided into various kinds of areas
  • area dividing means there is used a method whereby a discrimination regarding whether an area is a table or another area is made on the basis of a ratio of the area of circumscribed rectangle and the number of pixels, and the area is divided on the basis of the result of the discrimination.
  • an embodiment of the invention provides an image processing apparatus which can realize the reduction of the processing time when an area is divided, the decrease in size of a recording area, the improvement of a dividing precision of a complicated area construction, and the like.
  • an image processing apparatus as set out in claim 1.
  • the result is that without changing the existing area dividing process, (1) operating efficiency can be improved, (2) processing time can be reduced, (3) and the like.
  • Fig. 1 shows a schematic construction of an image processing apparatus of an embodiment of the invention.
  • reference numeral 101 denotes a unit for inputting image data
  • 102 a central processing unit (CPU) to control the apparatus and execute an arithmetic operation of a process in each unit
  • 103 a memory to store control programs of flowcharts, which will be explained hereinlater, and various kinds of data
  • 104 an output unit to generate the results of the operating processes, the results of the image processes, and the image data
  • 105 an area divider to execute a whole area dividing process.
  • reference numeral 106 denotes an image thinner to thin out an input image in a manner such that (m x n) pixels are thinned out by obtaining the OR of the vertical (m) and horizontal (n) dots of the input image and to form a single new pixel.
  • Reference numeral 107 denotes a labeler for labeling the pixel obtained by the thinning-out process and for simultaneously forming initial rectangle data. That is, a label is added to the black pixels of the image for every line and the same label is added to the pixels which are continuous in the vertical, lateral, or oblique direction and, at the same time, a rectangle is traced.
  • Reference numeral 108 denotes a separator detector to detect a separator, a table, a figure, or the like.
  • the separator detector 108 distinguishes a rectangle corresponding to a character portion (body), a rectangle corresponding to a figure or photograph, a rectangle corresponding to a table, a rectangle corresponding to a separator, and the like.
  • Reference numeral 109 denotes a writing direction detector to detect the writing direction of a sentence such as vertical writing, horizontal writing, or the like. The writing direction detector 109 presumes a vertical writing sentence or a horizontal writing sentence by comparing the width and height of the rectangle corresponding to the character portion (body).
  • Reference numeral 110 denotes an index detector to detect an index by using the writing direction and the character size.
  • Reference numeral 111 denotes a rectangle combiner to combine rectangles by using an imaginary rectangle which is obtained by enlarging the size of inherent rectangle.
  • FIG. 2 A flowchart of Fig. 2 shows an image processing procedure in the area divider 105 in Fig. 1.
  • a control program for executing the above image process has been stored in the memory 103.
  • An original image is inputted from the image input unit 101.
  • the OR of the vertical (m) dots and the horizontal (n) dots of the original image is calculated and the (m x n) pixels are thinned out, thereby newly forming one pixel.
  • the thinned-out pixel is set to black.
  • a label is added to the black pixel of the thinned-out pixel for every line and the same label is added to the pixels which are continuous in the vertical, lateral, or oblique direction and, at the same time, the rectangle is traced.
  • a label 1 is added to the pixel (A) which is first detected.
  • the coordinates (Xa, Ya) of the pixel A are set to an initial point and a terminal point of the rectangle, the number of pixels is set to 1, the same label 1 as that of the pixel is added to a rectangle label to distinguish the rectangle, and the above data is stored as rectangle data into the memory 103 as shown in Fig. 4.
  • the coordinates (Xb, Yb) of the pixel B are set to an initial point and a terminal point of the rectangle, the number of pixels is set to 1, the same label 2 as that of the pixel is added to the rectangle label to distinguish the rectangle, and the above data is also stored as rectangle data into the memory 103 as shown in Fig. 4.
  • the pixel label 1 is added to the pixel C. "1" is added to the number of pixels for the rectangle data of the rectangle label 1, so that the total number of pixels is equal to 2, the rectangle label is unchanged and is maintained to 1, and only the terminal point of the rectangle coordinates is updated from (Xa, Ya) to (Xa, Yc) (the coordinate of the initial point is unchanged).
  • the label 1 is added to the pixel D.
  • "1" is added to the number of pixels for the rectangle data of the rectangle label 1, so that the total number of pixels is set to 3.
  • the rectangle label is maintained to 1 without being changed. Only the terminal point of the rectangle coordinates is updated from (Xa, Yc) to (Xd, Yc) (the Y coordinate of the terminal point is unchanged).
  • both of the pixels D and B are obliquely continuous and are continuous with the pixel C, so that the label of the pixel B is changed from the label 2 to the label 1.
  • the number of pixels of the rectangle label 2 is added to that of the rectangle data of the rectangle label 1, so that the total number of pixels is set to 4.
  • the rectangle label is maintained to 1 without being changed. Only the terminal point of the rectangle coordinates is updated from (Xd, Yc) to (Xb, Yd) so as to include all of the pixels A, B, C, and D. With respect to the rectangle data of the rectangle label 2, it is invalidated by setting the rectangle label into 0.
  • the pixel label 1 is added to the pixel E. "1" is added to the number of pixels for the rectangle data of the rectangle label 1, so that the total number of pixels is set to 5.
  • the rectangle label is maintained to 1 without being changed.
  • the initial point is updated from (Xa, Ya) to (Xe, Ya) and the terminal point is updated from (Xb, Yd) to (Xb, Ye).
  • the rectangle label is set to 1
  • the initial point coordinates are set to (Xe, Ya)
  • the terminal point coordinates are set to (Xb, Ye)
  • the pixel label is set to 1
  • the number of pixels is set to 5.
  • the labeling is executed to all of the thinned-out pixels and the tracing of the rectangle is performed.
  • the separator detector 108 distinguishes a rectangle corresponding to the character portion (body), a rectangle corresponding to the figure, photograph, table, or the like, a rectangle corresponding to the separator, and the like by using the width W of the rectangle, height H, area S, and the number of pixels for the area, namely, the pixel density D (which can be easily obtained by calculating using those rectangle data).
  • the separator when the width W is equal to or less than a threshold value Tw1 and the height H is equal to or larger than the value which is twice as large as a threshold value Tw2 of the width W (step S503) or when the width W is larger than the threshold value Tw1 and the height H is equal to or larger than the value which is three times as large as a threshold value Tw3 of the width W (step S504), the separator is regarded as a separator which is vertically long and is unified by setting the rectangle label in the memory 103 into -3 and the pixel label constructing the rectangle keeps the current label number separately from the rectangle label (step S514).
  • a rectangle in which the above width and height were replaced is also judged in a manner similar to the above (steps S505, S506). If YES in each of the steps S505 and S506, the separator is regarded as a separator which is horizontally long and the rectangle label is changed to -3.
  • the separator is regarded as a modified separator such as a key-shape or the like and the rectangle label is set to -3 (step S514).
  • step S507 If NO in step S507 and when the area S is larger than a threshold value Ts1 (step S508), so long as the pixel density D is less than a threshold value Td2 (step S509), the rectangle is regarded as a table and the rectangle label is changed to -4 (step S515).
  • the rectangle is regarded as a figure or photograph and the rectangle label is changed to -5 (step S516).
  • a rectangle in which the pixel density D is equal to or larger than a threshold value Td3 (step S511) when the area S is equal to or less than the threshold value Ts1 and is equal to or larger than a threshold value Ts2 (step S510) or a rectangle in which both of the width W and the height H are equal to or larger than a threshold value Tw4 and the pixel density D is equal to or larger than a threshold value Td5 (step S513) is also regarded as a figure or photograph and the rectangle label is changed to -5 (step S516).
  • a rectangle in which the pixel density D is less than a threshold value Td4 (step S512) when the area S is equal to or less than the threshold value Ts1 and is equal to or larger than the threshold value Ts2 (step S510) is regarded as a table and the rectangle label is changed to -4 (step S515).
  • a rectangle corresponding to a figure, photograph, table, or the like, a rectangle corresponding to a separator, or the like is detected, the remaining rectangle is regarded as a body (character portion), and the rectangle label is unchanged and is set to the same label as the pixel label (step S517).
  • Fig. 6 shows the relation between the above threshold values in case of the area S and pixel density D and the classification of the separator or the like.
  • the writing direction detector 109 detects the writing direction of a sentence such as vertical writing, horizontal writing, or the like.
  • the writing direction detecting process will now be described with reference to a flowchart of Fig. 7.
  • the average values (w and h) of the width W and height H of the rectangle of the body are calculated (step S701).
  • the average width w is larger than the average height h
  • the sentence is regarded as a sentence having a large amount of horizontal writing sentences and the average height h is set to a character size of one character (steps S702, S703).
  • the average height h is larger than the average width w
  • the sentence is regarded as a sentence having a large amount of vertical writing sentences and the average width w is set to a character size of one character (steps S702, S710).
  • a histogram of rectangles is obtained in the direction opposite to the writing direction (steps S704, S711).
  • a position below a threshold value Tk is set to a separation of a paragraph from the shape of a peripheral distribution (steps S705, S712).
  • a histogram of the rectangles is obtained every paragraph in the same direction as the writing direction (steps S708, S715).
  • a length of continuous black pixels is set to a character size within the paragraph from the shape of the peripheral distribution and a length of continuous white pixels is detected as a line pitch (steps S709, S716). The above detecting process is repeated until the end of the paragraph (steps S706, S713).
  • an index is detected from the writing direction and the character size.
  • a rectangle whose width W is equal to or larger than the value of a threshold value Tml X the character size w is detected as a candidate of the index rectangle from the body rectangle (step S905).
  • a threshold value Tm2 the value of (a threshold value Tm2) times of the character size w
  • a sentence is regarded as a sentence in which the horizontal written characters are continuous and such a sentence is regarded as an index (step S908).
  • step S907 Since there is a case where a rectangle in which characters of a body were combined is also included in the rectangles which were regarded as indices from the size of rectangle as mentioned above, with respect to the rectangle in which the distance between the body rectangle and the index rectangle is closer than a threshold value T1 (step S907), it is corrected from the index to the body (step S909).
  • a rectangle such that none of the rectangles of the body and index exists within a range of the character size w is set into an index as an isolated rectangle (steps S910 to S913).
  • the rectangle combiner 111 the rectangles corresponding to the bodies which exist at random without any relation and the rectangles of the indices are combined.
  • FIG. 10 An example is shown in Fig. 10.
  • the combination of rectangles A1 and B will now be considered.
  • An imaginary rectangle A1' which is obtained by enlarging the rectangle A1 by only Px in the X direction and by only Py in the Y direction is now considered.
  • the periphery of the rectangle A1' is searched to see if there is a rectangle which is come into contact with or is inscribed in the rectangle A1' or not.
  • a rectangle which is in contact with A1' exists like a rectangle B
  • a combined rectangle A2 is newly formed.
  • the updated data of the rectangle A1 the updated data is stored and the data of the rectangle B is invalidated.
  • the values of Px and Py are obtained from the character size and line pitch.
  • the combining process itself is invalidated and the original rectangles are maintained as they are.
  • the indices are first combined and the body rectangles are subsequently combined.
  • the combining process itself is invalidated and the original rectangles are maintained as they are.
  • rectangle data of various kinds of areas obtained as mentioned above is generated from the output unit 104 to the outside together with the image data.
  • the area dividing process can be also executed even when the input image is a multivalue image such as a color image or the like.
  • step S202 when the number of pixels of the input image is sufficiently small to be a value which doesn't decrease the processing speed, such an image thinning process can be also omitted.
  • step S204 by distinguishing the rectangle label in dependence on a difference between the vertical and lateral directions of the separator, a difference between the threshold values when obtaining a figure, table, or the like, etc., the further fine area attributions can be classified.
  • step S204 in the case where only an image of a special attribute is inputted from the beginning like only a sentence or the like, the detecting process of a separator or the like can be also omitted.
  • step S205 in the detection of the writing direction in step S205, by again confirming the writing direction every paragraph, the processes can be also accurately executed to an original in which the vertical writing sentences and the horizontal writing sentences mixedly exist.
  • step S205 In the detection of the writing direction in step S205, when only the sentences in a predetermined writing direction are inputted from the beginning, by setting the writing direction to a fixed value, such a detecting process can be replaced.
  • step S206 when only the image having no index is inputted from the beginning, such a detecting process can be also omitted.
  • step S207 when a rectangle is come into contact with a rectangle such as a separator or the like, a check is further made to see if the rectangle is also in contact with the actually thinned-out pixel or not, and when the rectangle is not in contact with such a pixel, by executing the combining process, the apparatus can also cope with a slanted image.
  • step S208 only the image data of a necessary area can be also generated with reference to the rectangle data of various kinds of areas, so that the memory area and the processing time can be further reduced.
  • Fig. 12 is a block diagram showing an image processing apparatus of the embodiment.
  • reference numeral 101 denotes the input unit of image data.
  • Reference numeral 102 denotes the central processing unit (CPU) to control the apparatus and execute an arithmetic operation of the process in each unit in accordance with the control programs stored in the memory 103.
  • CPU central processing unit
  • Reference numeral 103 denotes the memory to store the control programs as shown in flowcharts, which will be explained hereinlater, and various kinds of data.
  • the memory 103 includes an ROM and an RAM.
  • Reference numeral 104 denotes the output unit to generates the results of the arithmetic operating processes, the results of the image processes, and image data.
  • Reference numeral 105 denotes the area divider showing the whole area dividing process.
  • Reference numeral 107 denotes the labeler for adding a label to the pixels of the input image and for simultaneously forming initial rectangle data.
  • Reference numeral 207 denotes an attribute detector to detect the attribute of a separator, a table, a figure, or the like.
  • Reference numeral 208 denotes a table rule detector to detect straight lines constructing a table.
  • Reference numeral 209 denotes a table area determiner to determine whether an area is a table area or not from a state of the table rules.
  • Fig. 13 is a flowchart showing the image process in the image processing apparatus of the embodiment.
  • the control programs to execute the image process have been stored in the memory 103.
  • the original image is first inputted from the image input unit 101.
  • the labels are added to the black pixels of the thinned-out image for every line and the same label is added to the pixels which are continuous in the vertical, lateral, and oblique directions, thereby simultaneously tracing a rectangle.
  • the label 1 is added to the pixel A which is first detected.
  • the coordinates (Xa, Ya) of the pixel A are set to an initial point and a terminal point of the rectangle, the number of pixels is set to 1, the same label is as that of the pixels is added to the rectangle label to distinguish the rectangle, and the above data is stored into the memory as rectangle data (Fig. 15).
  • the label 2 is added to the pixel B such that there is no continuous pixel in the left direction (since the pixel B is located on the first line, there is also no continuous pixel from the upper line).
  • the coordinates (Xb, Yb) of the pixel A are set to an initial point and a terminal point of the rectangle.
  • the number of pixels is set to 1 and the same label 2 as that of the pixels is added to the rectangle label to distinguish the rectangle.
  • the above data is also stored into the memory as rectangle data (Fig. 15).
  • the pixel label 1 Since the first pixel C of the second line is continuous with the pixel A of the label 1 from the upper line, the pixel label 1 is added to the pixel C. "1" is added to the number of pixels for the rectangle data of the rectangle label 1, so that the total number of pixels is set to 2. The rectangle label is maintained to 1 without being changed. Only the terminal point of the rectangle coordinates is updated from (Xa, Ya) to (Xa, Yc) (the coordinates of the initial point are unchanged).
  • the label 1 Since the next pixel D is continuous with the pixel C from the left side, the label 1 is added. "1" is added to the number of pixels for the rectangle data of the rectangle label 1, so that the total number of pixels is set to 3. The rectangle label is maintained to 1 without being changed. Only the terminal point of the rectangle coordinates is updated from (Xa, Yc) to (Xd, Yc) (the Y coordinate of the end point is unchanged).
  • the pixel D is also obliquely continuous with the pixel B and continuously comes from the pixel C, so that the label of the pixel B is changed from the label 2 to the label 1.
  • the number of pixels of the rectangle label 2 is added to that of the rectangle data of the rectangle label 1, so that the total number of pixels is set to 4.
  • the rectangle label is maintained to 1 without being changed. Only the terminal point of the rectangle coordinates is updated from (Xd, Yc) to (Xb, Yd) so as to include all of the pixels A, B, C, and D.
  • the rectangle data of the rectangle label 2 is invalidated by setting the rectangle label into 0.
  • the pixel label 1 Since the first pixel E of the third line is obliquely continuous with the pixel C, the pixel label 1 is added and "1" is added to the number of pixels for the rectangle data of the rectangle label 1, so that the total number of pixels is set to five.
  • the rectangle label is maintained to 1 without being changed.
  • the initial point of the rectangle coordinates is updated from (Xa, Ya) to (Xe, Ya) and the terminal point is updated from (Xb, Yd) to (Xb, Ye).
  • the labeling and tracing of the rectangle are executed for all of the thinned-out pixels.
  • a rectangle corresponding to the body after completion of the labeling and the tracing of the rectangle, a rectangle corresponding to the body, a rectangle corresponding to a figure, photograph, table, or the like, a rectangle corresponding to a separator, and the like are distinguished by using the width W of the rectangle, height H, area S, and the number of pixels for the area, namely, the pixel density D (those data can be easily obtained by calculating the rectangle data).
  • the separator is regarded as a separator which is vertically long and is unified by setting the rectangle label to -3, and the pixel labels constructing the rectangle is maintained as it is separately from the rectangle label (S514 in Fig. 16B).
  • the separator is regarded as a separator which is horizontally long and the rectangle label is changed to -3.
  • the separator is regarded as a modified separator such as key-shaped or the like and the rectangle label is changed to -3 (S514 in Fig. 16B).
  • a rectangle in which the pixel density D is less than a threshold value Td4 (S512 in Fig. 16B) is regarded as a table and the rectangle label is changed to -4 (S515 in Fig. 16B).
  • a width when calculating a histogram is decided from the width W and height H of the rectangle which was determined to be a table area (S901 in Fig. 20).
  • Partial histograms of the left and right edges in the vertical and horizontal directions are calculated (S902 in Fig. 20) for only the pixels (701 in Fig. 18, 801 in Fig. 19) (they can be easily obtained by comparing and selecting the pixel label of the rectangle data and the pixel label of each pixel in the rectangle) corresponding to the table frame excluding the bodies (702 in Fig. 18, 802 in Fig. 19) or the like in the rectangle which was determined to be a table area.
  • a threshold value Tkh of a histogram regarding whether a line is a table frame line which is parallel with the Y axis or not from the height of the rectangle and a threshold value Tkw of a histogram regarding whether a line is a table frame line which is parallel with the X axis or not from the width of rectangle are determined (S903, S904 in Fig. 20).
  • the corresponding rectangle is determined to be a table area (S905, S906, S908 in Fig. 20).
  • the corresponding rectangle is determined as a modified separator area (S905, S906, and S907 in Fig. 20).
  • step S201 by judging whether the input image is a multivalue image or not and converting the input image into the binary image, the area dividing process can be executed even when the input image is a multivalue image such as a color image.
  • the processing speed can be raised by the image t-hinning process for thinning out the image of (m dots in the vertical direction) x (n dots in the horizontal direction), thereby newly forming one pixel.
  • step S203 by distinguishing the rectangle labels by a difference between the vertical and horizontal directions of the separator, a difference between the threshold values when obtaining a figure, a table, or the like, etc., the further detailed area attributes can be also classified.
  • a difference Sx between the histograms at the left and right edges and a difference Sy between the histograms at the upper and lower edges are respectively obtained.
  • Ts a threshold value
  • the lengths of rules are not uniform, so that it is determined that the area is a modified separator area.
  • Ts a threshold value
  • step S206 only the image data of the necessary area can be also generated with reference to the rectangle data of various kinds of areas.
  • the memory area and the processing time can be further reduced.
  • histograms are calculated in the vertical and horizontal directions (S1501 in Fig. 26) for only the pixels (701 in Fig. 24, 801 in Fig. 25 ) (they can be easily obtained by comparing and selecting the pixel label of the rectangle data and the pixel label of each pixel in the rectangle) corresponding to the table frame excluding the bodies or the like (702 in Fig. 24, 802 in Fig. 25 ) in the rectangle which was determined to be a table area.
  • the threshold value Tkh of the histogram regarding whether the rules are parallel with the Y axis or not from the height of rectangle and the threshold value Tkw of the histogram regarding whether the rule is parallel with X axis or not from the width of rectangle are obtained (S1502 and S1503 in Fig. 26).
  • the position of the threshold value Tkh or more is regarded as a rule of the table area from the shape of the histogram in the vertical direction (S1504 in Fig. 26).
  • the position of the threshold value Tkw or more is similarly regarded as a rule of the table area from the shape of the histogram in the horizontal direction (S1505 in Fig. 26).
  • the number Nx of vertical rules obtained from the histogram in the vertical direction while regarding the portions of the threshold value Tkh or more as rules is stored.
  • the number Ny of horizontal rules obtained from the histogram in the horizontal direction while regarding the portions of the threshold value Tkw or more as rules is stored.
  • the area is determined to be a table area (S1506, S1508, S1509 in Fig. 26).
  • the area is decided to be a modified separator area (S1506, S1508, S1507 in Fig. 26).
  • rectangle data of various kinds of areas obtained as mentioned above is generated from the output unit 104 together with the image data.
  • step S201 by judging whether the input image is a multivalue image or not and by converting the input image into the binary image, the area dividing process can be performed even when the input image is a multivalue image such as a color image.
  • the processing speed can be also raised by the image thinning process for thinning out the image of (m dots in the vertical direction) x (n dots in the horizontal direction), thereby newly forming one pixel.
  • step S1204 when it is known that the inclination of the image after it was inputted is small, by tracing the lines in place of calculating a histogram, the table rules can be also detected.
  • step S1205 in place of fixing the threshold value TN to decide whether the area is a table area or not, as shown in Fig. 27, the area of rectangle is compared with a threshold value S1 and the preset number of rules is compared with threshold values Tn1 (area ⁇ S1) and Tn2 (area ⁇ S1).
  • Tn1 area ⁇ S1
  • Tn2 area ⁇ S1
  • the rules of the threshold values or more exist the area is determined to be a table area.
  • the number of rules is less than the threshold values, the area is decided as a modified separator area. Due to this, they can be replaced.
  • the threshold values which were determined in correspondence to the magnitude of the area of the rectangle which was temporarily determined as a table area are decided on the basis of the width and height of the rectangle and are respectively set to W1 and H1.
  • a threshold value of the number of vertical rules is set to TnH.
  • a threshold value of the number horizontal rules is set to TnW.
  • step S1206 only the image data of the necessary area can be also generated with reference to the rectangle data of the various kinds of areas.
  • the memory area and processing time can be further reduced.
  • the method of obtaining the histogram of the area whose attribute was determined to be a table area has been used as a method of determining the table area in the above description.
  • the table area is determined by the ratio of the area in the table frame to the area of the area which was decided to be a table area.
  • Fig. 33 is a flowchart for explaining in detail the determining process of the table area in step S204. The determining process will now be described hereinbelow with reference to the flowchart of Fig. 33.
  • step S221 an outline of an outer frame of a table is traced and point coordinates constructing the outline are stored.
  • Fig. 34 is a diagram for explaining the processes in steps S221 to S223.
  • reference numeral 2301 denotes a (very small) rectangle area which was determined to be a table; a white circle indicates a white pixel; a black dot denotes a black pixel; and a gray dot indicates a pixel which belongs to the inside of the table although it is a white pixel.
  • the black pixels or the like by characters in the table are omitted.
  • the inside within the table area is observed every line from the top and from the left side, thereby searching the black pixels constructing the table frame.
  • the black pixel is found out, its coordinates are stored into the memory 103.
  • the table area is observed in the direction from the coordinates (0, 1) to the coordinates (10, 1) to see if there is a black pixel exists or not every pixel. Since the black pixel exists at the position of the coordinates (2, 1), the coordinates are stored into the memory 103.
  • the outline is traced clockwise or counterclockwise from the black pixel and the coordinates of the black pixels (hereinafter, referred to as outline points) constructing the outline are stored into the memory 103.
  • outline points the coordinates up to (2, 1) - (10, 1), (10, 1) - (10, 11), (10, 11) - (1, 11), (0, 10) - (0, 4), and (1, 4) - (1, 2) are stored as outline points into the memory 103.
  • step S222 the number of pixels which belong to the inside of the table is counted every horizontal line and added.
  • two points at the rightmost edge and leftmost edge of the outline points stored in the memory 103 in step S221 are extracted for every line.
  • the distance between the two extracted points is added as the number of pixels belonging to the inside of the table.
  • the area in the table can be easily and rapidly obtained. Although such a process is sufficient as a determining process which is executed in the table area determiner 208, the area in the table can be also obviously accurately obtained.
  • step S223 a check is made to see if the numbers of pixels have been counted with respect to all of the lines in the table area or not. If YES, step S224 follows. When the line in which the number of pixels is not yet counted still remains, the processing routine is returned to step S222 and the processes are continued.
  • the coordinates of the outline point of the leftmost edge are set to (1, 2) and the coordinates of the outline point of the rightmost edge are set to (10, 2).
  • step S224 a ratio of the area S frame in the frame to the area S table of the rectangle area which was determined to be a table area is obtained.
  • step S225 a check is made to see if the ratio of the area in the frame is larger than a threshold value or not. Namely, when the following relation is satisfied, (area S frame in the frame) ⁇ (area S table of the rectangle area which was decided to be a table area) > (threshold value ⁇ ) the relevant rectangle is determined to be a table area in step S226.
  • the processing routine advances to step S205.
  • step S227 the relevant rectangle is decided to be a modified separator area and the rectangle label of the relevant rectangle stored in the memory 103 is changed to -3.
  • step S205 follows.
  • the area 2301 is determined to be a table area and step S205 follows.
  • step S205 follows.
  • reference numeral 701 denotes the rectangle area which was determined to be a table; 702 the frame; and 703 a line of a horizontal writing sentence.
  • the inside of the table surrounded by a frame 902 is shown like a black portion 1001 in Fig. 32 and the ratio of the area in the table frame is smaller than the threshold value ⁇ , so that such a portion is determined to be a modified separator.
  • reference numeral 901 denotes a rectangle area which was decided to be a table; 902 the frame; and 903 a line of a vertical writing sentence.
  • step S204 although the number of pixels belonging to the inside of the table has been counted every horizontal line and added in step S222, it is also possible to count and add the number of pixels belonging to the inside of the table for every vertical line.
  • step S204 although the number of pixels belonging to the inside of the table has been counted and added for every horizontal line in step S222, the table area determining process can be performed irrespective of the execution of the operation to count the number of black pixels constructing the sentence in the frame in this instance.
  • the table area determination in step S204 the number of pixels belonging to the inside of the table has been counted and added every horizontal line in step S222 and the area in the frame has been calculated.
  • the table area determining process can be executed even when the area in the frame is obtained by any other means.
  • step S205 only the image data of the necessary area can be also generated with reference to the rectangle data of various kinds of areas.
  • the memory area and the processing time can be further reduced.
  • Fig. 35 is a block diagram showing a construction of the image processing apparatus of the embodiment.
  • Reference numeral 117 denotes a display unit for overlapping the results divided by the area divider 105 to the original image data, for displaying a frame, and for displaying a list of attributes or the like.
  • Reference numeral 112 denotes an area division corrector to correct the erroneous division of the areas divided by the area divider 105.
  • Reference numeral 113 denotes an area attribute corrector to correct the area attribute which was presumed by the area divider 105.
  • Fig. 36 is a flowchart showing an area division correcting process in the image processing apparatus of the embodiment.
  • a control program to execute the area division correcting process has been stored in the memory 103.
  • Fig. 39 is a flowchart showing an area attribute correcting process in the image processing apparatus of the embodiment.
  • a control program to execute this process has been stored in the memory 103.
  • a plurality of areas which were erroneously divided by the area divider 105 are indicated by surrounding them by a frame (301 in Fig. 37) by using a pointing device or the like (S2501 in Fig. 36).
  • the corresponding area frames are shown by changing a display line from the solid lines to the dotted lines (S2503 in Fig. 36).
  • the designated frame is invalidated and erased and the processing routine is returned to the first step (S2502 in Fig. 36).
  • the new attributes are reversed and displayed (S2507 in Fig. 36).
  • the new attributes are determined (S2508 in Fig. 36)
  • either one of a plurality of areas existing in the designated frame is unconditionally selected and the coordinates, attributes, and the like of the area data are changed (S2509 in Fig. 36).
  • the other remaining area data is invalidated (S2510 in Fig. 36).
  • a new area frame (302 in Fig. 37) is displayed (S2511 in Fig. 36).
  • a plurality of areas are combined to one area by the above operation.
  • a target area (602 in Fig. 40) is designated by surrounding by a frame (603 in Fig. 40) by using a pointing device or the like (S2801 in Fig. 39) and a list table (601 in Fig. 40) of the attributes is displayed (S2802 in Fig. 39).
  • the current attributes are reversed and displayed (S2803 in Fig. 39) and the relevant area frame is changed from the solid line to the dotted line (602 in Fig. 40) (S2804 in Fig. 39).
  • the area in place of designating the area by using the pointing device or the like, the area can be also designated by vertically or horizontally moving a cursor by using a key on a keyboard.
  • the area data structure can include not only the data shown in the example of Fig. 38 but also data such as the number of pixels in the area, writing direction of the sentence, or the like.
  • the attribute corrector 113 in place of displaying the current attributes in the attribute list table by the reversing display, they can be also shown by changing the display color.
  • the attributes can be also changed.
  • Fig. 41 is a block diagram showing an image processing apparatus according to the third embodiment of the invention.
  • reference numeral 101 denotes the input unit of image data.
  • Reference numeral 102 denotes the central processing unit (CPU) to control the apparatus in accordance with the control programs stored in the memory 103 and to execute processing arithmetic operations in each section.
  • CPU central processing unit
  • Reference numeral 103 denotes a memory to store control programs and various kinds of data, which will be explained hereinlater, are stored.
  • Reference numeral 105 denotes the area divider to divide the area and to presume the attributes of the area.
  • Reference numeral 117 denotes the display unit for overlapping the results divided by the area divider 105 to the original data and displaying frames, thereby displaying a list table or the like of the attributes.
  • Reference numeral 112 denotes the division collector to correct the erroneous division of the areas divided by the area divider 105.
  • Reference numeral 113 denotes the attribute corrector to correct the area attributes which were presumed by the area divider 105.
  • Reference numeral 104 denotes the output unit to generate the results of the arithmetic operations, the results of the image processes, and image data.
  • Fig. 42 is a flowchart showing the area division correcting process in the image processing apparatus of the embodiment.
  • the control program to execute the above processes have been stored in the memory 103.
  • Fig. 45 is a flowchart showing an area attribute correcting process in the image processing apparatus of the embodiment.
  • the control program to execute this process has been stored in the memory 103.
  • one of a plurality of areas which were erroneously divided by the area divider 105 is designated and selected as a key area by the pointing device or the like (S201 in Fig. 42).
  • the area data (coordinates, attributes, etc.: refer to Fig. 44) of the key area is preserved in the memory 103 (S202 in Fig. 42).
  • Another divided area (refer to an area B in Fig. 43) is designated by the pointing device or the like (S203 in Fig. 42).
  • the area data (coordinates, attributes, etc.) of such an area is preserved into the memory 103 (S204 in Fig. 42).
  • a target area (refer to 602 in Fig. 46) is continuously designated twice by the pointing device or the like (S501 in Fig. 45) and the list table of the attributes (refer to 601 in Fig. 46) is displayed (S502 in Fig. 45).
  • the current attributes are reversed and displayed (S503 in Fig. 45) and the relevant area frame is displayed by changing from the solid line to the dotted line (refer to 602 in Fig. 46) (S504 in Fig. 45).
  • the area in place of designating the area by the pointing device or the like, the area can be also designated by vertically or horizontally moving a cursor by a key on a keyboard.
  • the area data structure can include not only the data shown as an example in Fig. 44 but also the data such as the number of pixels in the area, writing direction of the sentence, and the like.
  • the attribute corrector 113 in place of continuously designating the area twice by using the pointing device or the like, it is also possible to construct in a manner such that an attribute changing mode switch is provided and is turned on or off, thereby changing the meaning of the instruction of the pointing device or the like (specifically speaking, the ON-state of the attribute changing mode switch indicates the area instruction to change the attributes of the area, and the OFF-state indicates the area instruction to correct the division of the area).
  • the current attributes in the list table of the attributes can be also displayed by changing the display color in place of reversing and displaying.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Geometry (AREA)
  • Artificial Intelligence (AREA)
  • Multimedia (AREA)
  • Image Analysis (AREA)
  • Character Input (AREA)

Claims (23)

  1. Bildverarbeitungsvorrichtung, mit:
    einer Bereichsunterteilungseinrichtung zum Unterteilen von Bildinformationen in eine Vielzahl von Bildbereichen und automatischen Klassifizieren jedes unterteilten Bildbereichs als einen Zeichenbereich mit einer Vielzahl von Zeichen, einen Tabellenbereich, einen Trennungsbereich, oder einen Bereich entsprechend einer Figur oder einer Photographie und Zuweisen eines Attributs zu jedem unterteilten Bildbereich;
    einer Einrichtung zur Ausgabe von Daten zur Anzeige von Bildbereichen (602) zusammen mit den Eingabebildinformationen, wobei die die Vielzahl von Bereichen repräsentierenden Bildbereiche von der Bereichsunterteilungseinrichtung unterteilt sind;
    einer Einrichtung zum Empfang von Daten zur Angabe eines ausgewählten Bereichsbilds der Vielzahl von Bereichsbilder;
    einer Einrichtung (S2802, 2803) zur Ausgabe von Daten zur Anzeige einer Attributeliste (601), die die Vielzahl von Attributen aufweist, die einem unterteilten Bereich zugewiesen sein können, wobei das gegenwärtig zugewiesene Attribut des ausgewählten Bereichsbilds von anderen Attributen in der Attributeliste unterscheidbar ist;
    einer Einrichtung (S2805) zum Empfang von Daten zur Angabe eines bezeichneten Attributs unter der Vielzahl von in der angezeigten Attributeliste gezeigten Attribute; und
    einer Korrekturanweisungseinrichtung (S2807) zur Korrektur des Attributs des ausgewählten Bereichsbilds in das bezeichnete Attribut.
  2. Vorrichtung nach Anspruch 1, wobei die Bilderzeugungseinrichtung dahingehend ausgestaltet ist, um jeden Bildbereich innerhalb einer Umrissbox anzuzeigen.
  3. Vorrichtung nach Anspruch 2, wobei, wenn von der Auswahleinrichtung ein Bildbereich ausgewählt ist, die Bilderzeugungseinrichtung dahingehend ausgestaltet ist, um die das ausgewählte Bild umgebende Umrissbox auf eine Weise zu erzeugen, dass es von den anderen Bildbereichen unterscheidbar ist.
  4. Vorrichtung nach einem der Ansprüche 1 oder 3, wobei die Bereichsunterteilungseinrichtung umfasst:
    eine Bildausdünneinrichtung (106) zum ODER-en von m Punkten in der vertikalen Richtung mit n Punkten in der horizontalen Richtung der Eingabebildinformationen, um ausgedünnte Bilddaten zu erzeugen;
    eine Kennzeichnungseinrichtung (107) zum Hinzufügen eines Kennzeichens zu jeder Zeile von schwarzen Bildelementen der ausgedünnten Bildinformationen, Hinzufügen desselben Kennzeichens zu Bildelementen, welche in der vertikalen, horizontalen oder schrägen Richtung kontinuierlich sind, um so in der Lage zu sein, ein Rechteck aufzuspüren; und
    eine Erfassungseinrichtung (108) zum Unterscheiden eines Rechtecks entsprechend eines Satzabschnitts, eines Rechtecks entsprechend einer Figur oder Photographie, eines Rechtecks entsprechend einer Tabelle, eines Rechtecks entsprechend einer Trennung, und dergleichen, indem die Breite, die Höhe und der Bereich des Rechtecks zusammen mit dem von der Kennzeichnungseinrichtung hinzugefügten Kennzeichen verwendet wird, und indem die Bildelementdichte des Bereichs verwendet wird,
       wobei die eingegebenen Bildinformationen auf der Grundlage des Ergebnisses der Erfassung der Erfassungseinrichtung in die Bildbereiche unterteilt werden.
  5. Vorrichtung nach einem der voranstehenden Ansprüche, wobei die Bereichsunterteilungseinrichtung zudem umfasst:
    eine Schreibrichtungs-Erfassungseinrichtung (109) zum Vornehmen einer Annahme darüber, ob ein von der Erfassungseinrichtung erfasster Satzabschnitt vertikal oder horizontal ist, auf der Grundlage der Breite und der Höhe des Rechtecks entsprechend dem Satzabschnitt;
    eine Indexerfassungseinrichtung (110) zur Erfassung eines Indexes unter Verwendung der Schreibrichtung, die von der Schreibrichtungs-Erfassungseinrichtung entschieden ist, für eine Zeichengröße; und
    eine Rechteck- (102) Kombinationseinrichtung zur Kombination der Rechtecke unter Verwendung eines imaginären Rechtecks, das durch Vergrößern der Größe eines ausgewählten Rechtecks erlangt wird.
  6. Vorrichtung nach einem der Ansprüche 1 bis 3, wobei die Bereichsunterteilungseinrichtung umfasst:
    eine Rechteckerkennungseinrichtung zur Erkennung eines Rechtecks, indem kontinuierliche schwarze Bildelemente der Bildinformationen erfasst werden;
    eine Attributunterscheidungseinrichtung zur Unterscheidung von Attributen des Rechtecks unter Verwendung der Breite, der Höhe, des Bereichs und der Bildelementdichte des von dem erkannten Rechteck angegebenen Bereichs; und
    eine Steuereinrichtung, die dahingehend ausgestaltet ist, um Attribute jedes erkannten Bereichs durch Berechnung und Vergleich von Teilhistogrammen des Bereichs zu bestimmen.
  7. Vorrichtung nach einem der Ansprüche 1 bis 3, wobei die Bereichsunterteilungseinrichtung umfasst:
    eine Rechteckerkennungseinrichtung zur Erkennung eines Rechtecks, indem kontinuierliche schwarze Bildelemente der Bildinformationen erfasst werden;
    eine Attributunterscheidungseinrichtung zur Unterscheidung von Attributen des erkannten Rechtecks unter Verwendung der Breite, der Höhe, des Bereichs und der Bildelementdichte eines von dem Rechteck angegebenen Bereichs;
    eine Maßstaberfassungseinrichtung zur Erfassung von Maßstäben in dem erkannten Bereich; und
    eine Steuereinrichtung zur Steuerung der Bestimmungsattribute des Bereichs durch die erfasste Anzahl der Maßstäbe.
  8. Vorrichtung nach einem der Ansprüche 1 bis 3, wobei die Bereichsunterteilungseinrichtung umfasst:
    eine Rechteckerkennungseinrichtung zur Erkennung eines Rechtecks, indem kontinuierliche schwarze Bildelemente der Bildinformationen erfasst werden;
    eine Attributunterscheidungseinrichtung zur Unterscheidung von Attributen des erkannten Rechtecks unter Verwendung der Breite, der Höhe, des Bereichs und der Bildelementdichte eines von dem Rechteck angegebenen Bereichs; und
    eine Steuereinrichtung, die dahingehend ausgestaltet ist, um Attribute des Bereichs aus einem Verhältnis eines Bereichs des Bereichs, welcher als ein Tabellenbereich und ein Bereich in einem Tabellenrahmen entschieden wurde, zu bestimmen.
  9. Vorrichtung nach einem der Ansprüche 6, 7 oder 8, wobei die Attribute entweder einen Körper, eine Figur oder eine Tabelle angeben.
  10. Vorrichtung nach einem der Ansprüche 1 bis 3, wobei die Korrekturanweisungseinrichtung dahingehend ausgestaltet ist, um spezielle Unterteilungsbereiche zu bezeichnen, die unter einer Vielzahl von Unterteilungsbereichen zu koppeln sind, die in der eingegebenen Bildinformationen enthalten sind; und
       die Aktualisiersteuereinrichtung dahingehend ausgestaltet ist, um die gespeicherten Attributinformationen gemäß der Kopplung der Unterteilungsbereiche zu aktualisieren, wenn sie wie von der Korrekturanweisungseinrichtung bezeichnet korrigiert sind, um so die original unterteilten Bildbereiche zu korrigieren.
  11. Vorrichtung nach Anspruch 9, zudem mit einer Einrichtung zur Korrektur von Attributen der bezeichneten Bereiche im Zusammenhang mit der Unterteilung der Bereiche.
  12. Bildverarbeitungsverfahren, mit den Schritten des:
    Unterteilens von Bildinformationen in eine Vielzahl von Bildbereichen und automatischen Klassifizierens jedes unterteilten Bildbereichs als einen Zeichenbereich mit einer Vielzahl von Zeichen, einen Tabellenbereich, einen Trennungsbereich, oder einen Bereich entsprechend einer Figur oder einer Photographie und Zuweisen eines Attributs zu jedem unterteilten Bildbereich;
    Erzeugens einer Anzeige von Bildbereichen (602) zusammen mit den Eingabebildinformationen, wobei die die Vielzahl von Bereichen repräsentierenden Bildbereiche von der Bereichsunterteilungseinrichtung unterteilt sind;
    Steuerns der Auswahl eines gewünschten Bereichsbildes der Vielzahl von angezeigten Bereichsbilder;
    Erzeugens von Anzeigedaten zur Anzeige einer Attributeliste (601), die die Vielzahl von Attributen aufweist, welche einem unterteilten Bereich zugewiesen sein können, wobei das gegenwärtig zugewiesene Attribut des ausgewählten Bereichsbilds von anderen Attributen in der Attributeliste unterscheidbar ist;
    Angebens eines bezeichneten Attributs unter der Vielzahl von in der angezeigten Attributeliste gezeigten Attribute; und
    Korrigierens des Attributs des ausgewählten Bereichsbilds in das bezeichnete Attribut.
  13. Verfahren nach Anspruch 12, wobei die angezeigten Bildbereiche jeweils in einer Umrissbox eingeschlossen sind.
  14. Verfahren nach Anspruch 13, wobei, wenn ein Bildbereich ausgewählt ist, die den ausgewählten Bildbereich umgebende Umrissbox von den anderen Umrissboxen unterscheidbar ist.
  15. Verfahren nach einem der Ansprüche 12 bis 14, wobei der Bereichsunterteilungsschritt die Schritte umfasst:
    ODER-en von m Punkten des Eingabebildes der Eingabebildinformationen in der vertikalen Richtung mit n Punkten in der horizontalen Richtung, um ausgedünnte Bilddaten zu erzeugen;
    Hinzufügen eines Kennzeichens zu jeder Zeile von schwarzen Bildelementen der ausgedünnten Bildinformationen, Hinzufügen desselben Kennzeichens zu den Bildelementen, welche in der vertikalen, horizontalen oder schrägen Richtung kontinuierlich sind, und gleichzeitig Aufspüren eines Rechtecks; und
    Unterscheiden eines Rechtecks entsprechend einem Satzabschnitt, eines Rechtecks entsprechend einer Figur oder Photographie, eines Rechteck entsprechend einer Tabelle, eines Rechtecks entsprechend einer Trennung, und dergleichen, indem die Breite, die Höhe und der Bereich des Rechtecks verwendet wird, zu dem das Kennzeichen durch die Kennzeichnungseinrichtung hinzugefügt ist, und indem die Anzahl von Bildelementen für den Bereich, nämlich eine Bildelementdichte, verwendet wird,
       wobei ein Bereich der eingegebenen Bildinformationen auf der Grundlage des Ergebnisses der Erfassung der Erfassungseinrichtung unterteilt wird.
  16. Verfahren nach einem der Ansprüche 12 bis 14, wobei der Bereichsunterteilungsschritt zudem umfasst:
    Annehmen eines vertikalen Schreibsatzes oder eines horizontalen Schreibsatzes durch Vergleichen der Breite und der Höhe des Rechtecks entsprechend einem Satzabschnitt, der von der Erfassungseinrichtung unterschieden ist;
    Erfassen eines Indexes unter Verwendung der Schreibrichtung, die von der Schreibrichtungs-Erfassungseinrichtung entschieden ist, für eine Zeichengröße; und
    Kombinieren der Rechtecke unter Verwendung eines imaginären Rechtecks, das durch Vergrößern der Größe eines ausgewählten Rechtecks erlangt wird.
  17. Verfahren nach einem der Ansprüche 12 bis 14, wobei der Bereichsunterteilungsschritt umfasst:
    Erkennen eines Rechtecks, indem kontinuierliche schwarze Bildelemente erfasst werden;
    Unterscheiden von Attributen des Rechtecks unter Verwendung der Breite, der Höhe, des Bereichs und der Bildelementdichte eines von dem erkannten Rechteck angegebenen Bereichs; und
    Bestimmen von Attributen des erkannten Bereichs durch Berechnen und Vergleichen von Teilhistogrammen des Bereichs.
  18. Verfahren nach einem der Ansprüche 12 bis 14, wobei der Bereichsunterteilungsschritt umfasst:
    Erkennen eines Rechtecks, indem kontinuierliche schwarze Bildelemente der Bildinformationen erfasst werden;
    Unterscheiden von Attributen des erkannten Rechtecks unter Verwendung der Breite, der Höhe, des Bereichs und der Bildelementdichte eines von dem Rechteck angegebenen Bereichs;
    Erfassen von Maßstäben in dem erkannten Bereich; und
    Bestimmen von Attributen des Bereichs durch die Anzahl der Maßstäbe.
  19. Verfahren nach einem der Ansprüche 12 bis 14, wobei der Bereichsunterteilungsschritt die Schritte umfasst:
    Erkennen eines Rechtecks, indem kontinuierliche schwarze Bildelemente der Bildinformationen erfasst werden;
    Unterscheiden von Attributen des erkannten Rechtecks unter Verwendung der Breite, der Höhe, des Bereichs und der Bildelementdichte eines von dem Rechteck angegebenen Bereichs; und
    Bestimmen von Attributen des Bereichs aus einem Verhältnis eines Bereichs des Bereichs, welcher als ein Tabellenbereich oder ein Bereich in einem Tabellenrahmen entschieden wurde.
  20. Verfahren nach einem der Ansprüche 15, 16 oder 17, wobei die Attribute entweder einen Körper, eine Figur oder eine Tabelle angeben.
  21. Verfahren nach einem der Ansprüche 12 bis 20, wobei der Korrekturanweisungsschritt spezielle Unterteilungsbereiche bezeichnet, die unter einer Vielzahl von Unterteilungsbereichen zu koppeln sind, die in den eingegebenen Bildinformationen erfasst sind; und
       der Aktualisiersteuerschritt die gespeicherten Attributinformationen gemäß eines Zusammenkoppelns der Unterteilungsbereiche, die von der Korrekturanweisungseinrichtung bezeichnet sind, und eines Korrigierens der Originalenunterteilungsbereiche aktualisiert.
  22. Verfahren nach Anspruch 21, zudem mit dem Schritt des Korrigierens von Attributen der bezeichneten Bereiche im Zusammenhang mit der Unterteilung der Bereiche.
  23. Speichermedium, welches von einem Prozessor ausführbare Anweisungen zur Steuerung eines Prozessors speichert, um alle die Schritte des Verfahrens nach einem der Ansprüche 12 bis 22 auszuführen.
EP93307181A 1992-09-11 1993-09-10 Verfahren und Anordnung zur Bildverarbeitung Expired - Lifetime EP0587450B1 (de)

Applications Claiming Priority (9)

Application Number Priority Date Filing Date Title
JP243252/92 1992-09-11
JP4243252A JPH0696275A (ja) 1992-09-11 1992-09-11 画像処理装置
JP24325292 1992-09-11
JP256315/92 1992-09-25
JP4256315A JPH06110994A (ja) 1992-09-25 1992-09-25 画像処理装置
JP25631592 1992-09-25
JP2700/93 1993-01-11
JP270093 1993-01-11
JP5002700A JPH06208625A (ja) 1993-01-11 1993-01-11 画像処理方法及び装置

Publications (2)

Publication Number Publication Date
EP0587450A1 EP0587450A1 (de) 1994-03-16
EP0587450B1 true EP0587450B1 (de) 2004-11-17

Family

ID=27275487

Family Applications (1)

Application Number Title Priority Date Filing Date
EP93307181A Expired - Lifetime EP0587450B1 (de) 1992-09-11 1993-09-10 Verfahren und Anordnung zur Bildverarbeitung

Country Status (3)

Country Link
US (2) US6104832A (de)
EP (1) EP0587450B1 (de)
DE (1) DE69333694T2 (de)

Families Citing this family (37)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0587450B1 (de) * 1992-09-11 2004-11-17 Canon Kabushiki Kaisha Verfahren und Anordnung zur Bildverarbeitung
DE69531882D1 (de) * 1994-07-13 2003-11-13 Canon Kk Gerät und Verfahren zur Bildverarbeitung
JP3814320B2 (ja) * 1995-12-14 2006-08-30 キヤノン株式会社 画像処理方法及び装置
US6389162B2 (en) 1996-02-15 2002-05-14 Canon Kabushiki Kaisha Image processing apparatus and method and medium
JPH10173917A (ja) * 1996-10-11 1998-06-26 Nikon Corp 画像読取システム、画像読取装置及び制御手順を記憶する記憶媒体
US6351558B1 (en) * 1996-11-13 2002-02-26 Seiko Epson Corporation Image processing system, image processing method, and medium having an image processing control program recorded thereon
JP4035228B2 (ja) * 1998-05-11 2008-01-16 キヤノン株式会社 画像処理方法及び画像処理装置
EP0999522B1 (de) * 1998-11-06 2008-04-09 Canon Kabushiki Kaisha Bildverarbeitungsverfahren, -system und -gerät, und Speichermedium
US6643400B1 (en) 1999-03-31 2003-11-04 Minolta Co., Ltd. Image processing apparatus and method for recognizing specific pattern and recording medium having image processing program recorded thereon
JP3756719B2 (ja) 2000-01-20 2006-03-15 理想科学工業株式会社 文書修飾装置及び画像処理装置
JP3927776B2 (ja) * 2001-06-25 2007-06-13 キヤノン株式会社 画像処理装置及びその処理方法
JP3833138B2 (ja) 2002-04-19 2006-10-11 キヤノン株式会社 画像処理装置、画像処理方法、プログラム及び記憶媒体
JP4165706B2 (ja) * 2004-01-26 2008-10-15 株式会社リコー 原稿読み取り装置および画像形成装置
US7565034B2 (en) * 2004-06-17 2009-07-21 Avago Technologies Ecbu Ip (Singapore) Pte. Ltd. Determination of a navigation window in an optical navigation system
US20060263133A1 (en) * 2005-05-17 2006-11-23 Engle Jesse C Network based method and apparatus for collaborative design
JP4240107B2 (ja) * 2006-10-27 2009-03-18 コニカミノルタビジネステクノロジーズ株式会社 領域判定方法、領域判定装置、画像処理装置、およびコンピュータプログラム
US8144989B2 (en) * 2007-06-21 2012-03-27 Sharp Laboratories Of America, Inc. Methods and systems for identifying text orientation in a digital image
US8208725B2 (en) * 2007-06-21 2012-06-26 Sharp Laboratories Of America, Inc. Methods and systems for identifying text orientation in a digital image
US8340430B2 (en) * 2007-07-10 2012-12-25 Sharp Laboratories Of America, Inc. Methods and systems for identifying digital image characteristics
US8023741B2 (en) * 2008-05-23 2011-09-20 Sharp Laboratories Of America, Inc. Methods and systems for detecting numerals in a digital image
US8023770B2 (en) * 2008-05-23 2011-09-20 Sharp Laboratories Of America, Inc. Methods and systems for identifying the orientation of a digital image
US8160365B2 (en) * 2008-06-30 2012-04-17 Sharp Laboratories Of America, Inc. Methods and systems for identifying digital image characteristics
US8620080B2 (en) * 2008-09-26 2013-12-31 Sharp Laboratories Of America, Inc. Methods and systems for locating text in a digital image
JP4706764B2 (ja) * 2009-01-29 2011-06-22 富士ゼロックス株式会社 画像処理装置及び画像処理プログラム
JP5842441B2 (ja) 2011-07-29 2016-01-13 ブラザー工業株式会社 画像処理装置およびプログラム
JP5853470B2 (ja) 2011-07-29 2016-02-09 ブラザー工業株式会社 画像処理装置、画像処理プラグラム
JP5796392B2 (ja) 2011-07-29 2015-10-21 ブラザー工業株式会社 画像処理装置、および、コンピュータプラグラム
JP5776419B2 (ja) 2011-07-29 2015-09-09 ブラザー工業株式会社 画像処理装置、画像処理プラグラム
JP5801665B2 (ja) 2011-09-15 2015-10-28 キヤノン株式会社 固体撮像装置、ａ／ｄ変換器およびその制御方法
JP5806566B2 (ja) 2011-09-15 2015-11-10 キヤノン株式会社 Ａ／ｄ変換器および固体撮像装置
JP5974589B2 (ja) * 2012-03-30 2016-08-23 ブラザー工業株式会社 画像処理装置およびプログラム
US9076058B2 (en) 2013-01-29 2015-07-07 Sharp Laboratories Of America, Inc. Methods, systems and apparatus for determining orientation in a document image
US9058539B2 (en) 2013-04-16 2015-06-16 Canon Kabushiki Kaisha Systems and methods for quantifying graphics or text in an image
US9310983B2 (en) * 2013-10-16 2016-04-12 3M Innovative Properties Company Adding, deleting digital notes from a group of digital notes
WO2015057804A1 (en) 2013-10-16 2015-04-23 3M Innovative Properties Company Organizing digital notes on a user interface
CN106940799B (zh) * 2016-01-05 2020-07-24 腾讯科技（深圳）有限公司 文本图像处理方法和装置
CN108363879A (zh) * 2018-02-27 2018-08-03 杭州深绘智能科技有限公司 适用于服装图像的数据处理方法

Family Cites Families (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH0738681B2 (ja) * 1987-03-20 1995-04-26 富士ゼロックス株式会社 領域認識装置
US4926251A (en) * 1987-04-07 1990-05-15 Kabushiki Kaisha Toshiba Color image processing apparatus with image corrector
KR920009343B1 (ko) * 1987-05-15 1992-10-15 후지 제록스 가부시끼가이샤 칼라마킹장치
JPH0821057B2 (ja) * 1987-07-10 1996-03-04 日本電気株式会社 文書画像解析方式
US4856075A (en) * 1987-10-05 1989-08-08 Eastman Kodak Company Image discrimination
US5113251A (en) * 1989-02-23 1992-05-12 Fuji Xerox Co. Editing control system and area editing system for image processing equipment
JP2940936B2 (ja) * 1989-06-06 1999-08-25 株式会社リコー 表領域識別方法
US5202933A (en) * 1989-12-08 1993-04-13 Xerox Corporation Segmentation of text and graphics
JPH03214967A (ja) * 1990-01-19 1991-09-20 Minolta Camera Co Ltd 画像読取り装置
JPH04248687A (ja) * 1991-01-23 1992-09-04 Internatl Business Mach Corp <Ibm> 文書画像のレイアウト解析方法及びシステム
US5703962A (en) 1991-08-29 1997-12-30 Canon Kabushiki Kaisha Image processing method and apparatus
US5384863A (en) * 1991-11-19 1995-01-24 Xerox Corporation Methods and apparatus for automatic modification of semantically significant portions of a document without document image decoding
DE69331126T2 (de) 1992-07-31 2002-06-20 Canon Kk Bildverarbeitungsgerät und Verfahren zur mehrstufigen Kompression
EP0587450B1 (de) * 1992-09-11 2004-11-17 Canon Kabushiki Kaisha Verfahren und Anordnung zur Bildverarbeitung
JP3214967B2 (ja) 1993-12-24 2001-10-02 東北リコー株式会社 孔版式製版印刷装置及び排版ロールの巻取軸

Also Published As

Publication number Publication date
US6385338B1 (en) 2002-05-07
EP0587450A1 (de) 1994-03-16
US6104832A (en) 2000-08-15
DE69333694T2 (de) 2005-10-20
DE69333694D1 (de) 2004-12-23

Similar Documents

Publication Publication Date Title
EP0587450B1 (de) Verfahren und Anordnung zur Bildverarbeitung
US7627148B2 (en) Image data processing apparatus and method, and image data processing program
US5572602A (en) Image extraction system for extracting patterns such as characters, graphics and symbols from image having frame formed by straight line portions
EP0854434B1 (de) Gerät und Verfahren zur Extrahierung von Tabellengeraden innerhalb normaler Dokumentbilder
US7536050B2 (en) Using graph cuts for editing photographs
EP1091320A2 (de) Verarbeitung von mehreren numerischen Bildern
EP0935216A2 (de) Farbbildverarbeitungsgerät und Gerät zum Extrahieren von Mustern
EP1081648B1 (de) Verfahren zum Verarbeiten eines numerischen Bildes
EP1956551B1 (de) Mustererkennungsvorrichtung, mustererkennungsverfahren und mustererkennungsprogramm
US20080069398A1 (en) Code image processing method
EP0381773B1 (de) Buchstabenerkennungsvorrichtung
US6937762B2 (en) Image processing device and program product
US6005976A (en) Image extraction system for extracting patterns such as characters, graphics and symbols from image having frame formed by straight line portions
EP0779593B1 (de) Verfahren und Gerät zur Bildverarbeitung und Speichermedium
US6493470B1 (en) Image processing method and apparatus for detecting the tilt amount of input image data
US5625710A (en) Character recognition apparatus using modification of a characteristic quantity
EP0898240B1 (de) Gerät und Verfahren zur Bildverarbeitung
US7386172B2 (en) Image recognition method
JPH06208625A (ja) 画像処理方法及び装置
CN102682308B (zh) 图像处理方法和图像处理设备
CN109961063B (zh) 文本检测方法及装置、计算机设备和存储介质
JP2868134B2 (ja) 画像処理方法及び装置
JPH02293989A (ja) 文字認識装置
JP2827960B2 (ja) 宛名行抽出装置
JP2000207490A (ja) 文字切出装置、および文字切出方法

Legal Events

Date Code Title Description
PUAI Public reference made under article 153(3) epc to a published international application that has entered the european phase

Free format text: ORIGINAL CODE: 0009012

AK Designated contracting states

Kind code of ref document: A1

Designated state(s): DE FR GB

17P Request for examination filed

Effective date: 19940801

17Q First examination report despatched

Effective date: 19991007

GRAP Despatch of communication of intention to grant a patent

Free format text: ORIGINAL CODE: EPIDOSNIGR1

GRAS Grant fee paid

Free format text: ORIGINAL CODE: EPIDOSNIGR3

GRAA (expected) grant

Free format text: ORIGINAL CODE: 0009210

AK Designated contracting states

Kind code of ref document: B1

Designated state(s): DE FR GB

REG Reference to a national code

Ref country code: GB

Ref legal event code: FG4D

REF Corresponds to:

Ref document number: 69333694

Country of ref document: DE

Date of ref document: 20041223

Kind code of ref document: P

ET Fr: translation filed
PLBE No opposition filed within time limit

Free format text: ORIGINAL CODE: 0009261

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: NO OPPOSITION FILED WITHIN TIME LIMIT

26N No opposition filed

Effective date: 20050818

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: FR

Payment date: 20080923

Year of fee payment: 16

REG Reference to a national code

Ref country code: FR

Ref legal event code: ST

Effective date: 20100531

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: FR

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20090930

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: DE

Payment date: 20110930

Year of fee payment: 19

Ref country code: GB

Payment date: 20110922

Year of fee payment: 19

GBPC Gb: european patent ceased through non-payment of renewal fee

Effective date: 20120910

REG Reference to a national code

Ref country code: DE

Ref legal event code: R119

Ref document number: 69333694

Country of ref document: DE

Effective date: 20130403

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: GB

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20120910

Ref country code: DE

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20130403